Umami vs Plausible
Both Umami and Plausible are open source, privacy-focused web analytics platforms that run in Docker, collect visitor metrics without cookies, and position themselves as GDPR-compliant alternatives to Google Analytics. I ran both simultaneously on my personal sites to decide which to keep long-term. My conclusion was Umami, and it wasn’t particularly close once I moved past surface aesthetics. The deciding factors were practical: API flexibility, navigational coherence, and—counterintuitively—Plausible’s own setup flow working against it. Plausible is marginally prettier in places, but it squanders that advantage with some genuinely puzzling navigation decisions.
Setup and Deployment
Getting Umami running is straightforward: a single app container and a PostgreSQL database, a few lines of Docker Compose, and you’re at the dashboard in minutes. The resource footprint is light.
Plausible Community Edition is meaningfully more complex. The self-hosted stack requires PostgreSQL, ClickHouse, and the application container—three services instead of two, with ClickHouse bringing its own configuration and resource demands. ClickHouse is a columnar database optimized for analytical queries at scale, which is genuinely the right tool for the job, but it’s a significant operational dependency to take on for a personal site. Once running, both platforms are stable and reliable. But Plausible asks more of you to get there, and that gap doesn’t close.
The User Interface
Umami’s interface is fast, clean, and immediately comprehensible. The dashboard for each site shows the metrics you actually care about—pageviews, visitors, bounce rate, referrers, pages, devices—without ceremony. Navigation is logical throughout.
Plausible’s interface is marginally more polished visually. The typography is a little tighter, the color choices slightly more refined. If you’re judging screenshots, Plausible wins. If you’re actually using either product, the gap narrows fast.
Where Plausible loses the thread is during initial site setup. The onboarding flow has a strange navigational quality—it redirects you around in ways that don’t feel purposeful, the script installation step interrupts the flow oddly, and switching between sites once you have several involves more clicks than it should. None of this is catastrophic, but it accumulates. A tool you use every day should get out of its own way, and Plausible doesn’t always manage that.
Multi-Site Management
This is Umami’s one genuine weakness. Each site in Umami is fully isolated—its own data, its own view, no unified cross-site dashboard. To see traffic across multiple sites, you visit each one in turn. Conceptually, this separation makes sense: it supports distinct teams, data segregation, and independent access controls. For a larger organization, those properties matter.
For a small site admin with a handful of personal properties, the isolation is friction without benefit. Plausible handles multi-site navigation more naturally—the main listing gives you a quick read on all your sites before drilling into any one of them. This is the one area where I’d take Plausible’s approach over Umami’s without hesitation.
The API
This is where the decision became clear. Umami’s API accepts historical event data—you can send events with past timestamps and they land correctly in the analytics record. This may sound like an edge case until you encounter a situation where it isn’t.
The immediate use for me was GitHub traffic data. GitHub provides two weeks of referrer and pageview data through its API; Umami let me import that history rather than discard it. More broadly, any site with years of existing traffic history faces a hard choice when switching analytics platforms: start fresh and lose the record, or find a tool that lets you migrate. Umami’s API makes migration a real option. Plausible’s API is primarily read-oriented—it exposes your data for export and integration, but it doesn’t readily accept historical imports. If your sites are established and your traffic history matters, that asymmetry is significant.
The Verdict
Umami wins on the factors that matter operationally. The API is more capable, the navigation is more coherent, and the deployment footprint is lighter. Plausible has a slight aesthetic edge and handles multi-site browsing more cleanly, but those advantages don’t outweigh a more functional API and a setup experience that doesn’t fight you.
The one real limitation—no unified dashboard across sites—is something you learn to live with. It’s a conceptual decision, not an oversight, and the workaround is trivial: bookmark each site’s dashboard. For the API flexibility alone, Umami earns its place in the stack.

